What If We Each Had Our Own Laws?

The rise of Big Data could give us tools to create different rules for different people. In a new book, Professors Omri Ben-Shahar and Ariel Porat argue that personalized law might make us more equal.

Right now, posted speed limits apply to all drivers, regardless of individual experience and skill. In the United States, the wealthy and the poor pay the same parking fines, even though a $100 penalty is likely to have a greater effect on the latter. Individuals of differing abilities face the same minimum-age limits on activities like alcohol consumption, employment, and driving; consumer protections apply regardless of vulnerability and need; and borrowers receive—and often fail to read—the same long list of loan disclosures.

But what if laws weren’t quite so … uniform? What if the rules and regulations governing our lives were more like tailored suits, customized by algorithms to account for individual preferences, skills, experience, traits, needs, and behavior—and calibrated to reduce accidents, inequality, crime, unnecessary costs, information overload, and more? Would you want your own laws? And perhaps more importantly: Would the system be more effective, more efficient, and more equitable—or would it become a Big Brotherly invasion of privacy in which existing biases continue to reign, human judgment becomes relegated to machines, and savvy swindlers game the system?

cover of Personalized Law

These are the questions Professor Omri Ben-Shahar, the University of Chicago’s Leo and Eileen Herzel Professor of Law, and Ariel Porat, the president of Tel Aviv University and a longtime UChicago Law faculty visitor, ask in a new book aimed at jumpstarting a conversation that is both difficult and timely.

Prediction algorithms are already being used to aid in criminal sentencing, and simple computer-generated legal advice is available online. Personalization is, increasingly, a part of life, and given the fast-rising ability to gather, store, and interpret vast amounts of individual data, the ability to create customized laws could soon follow.

If we can navigate the pitfalls, the authors argue, we might finally create a legal system that “treats people as individuals not a population.”

“Uniformity and equality under the law is a myth,” said Ben-Shahar, a leading expert in contracts, consumer protections, and law and economics and the coauthor of Personalized Law: Different Rules for Different People (Oxford University Press, 2021). “When we talk about laws to protect the weak, we’re thinking about specific populations who are being treated unfairly or hurt in some way. But then we give the protections to everyone, and people who are more sophisticated and educated and affluent enjoy the protections disproportionately more. [The typical response then] is to say we need to restore uniformity. But we don’t need more uniformity, we need to embrace differentiated treatment—but in a way that's good rather than broken.”

In recent years, Ben-Shahar, Porat, and other legal scholars have begun to imagine the possibilities: personalized speed limits delivered directly to each driver, consumer protections that target those most in need, and tailored default rules designed to more accurately predict an individual’s preferences. In Porat’s earlier work with Professor Lior Strahilevitz, they envisioned a specific algorithm that might use collected data to predict an individual’s bequest wishes; should that individual die without a will, the default estate allocation would, in theory, more closely fit her desires. In the book, Ben-Shahar and Porat point to applications in many areas of the law. In criminal law, for example, algorithms are already used to predict which defendants pose greater risks, and such prediction could be used in sentencing, parole, and bail proceedings. Sanctions could be calibrated to reflect the likelihood of detection (with higher penalties when the probability of being caught is low) and the potential benefit to the offender (with higher penalties when the crime promises a bigger likely payoff). Sliding penalties could even level the impact among offenders by accounting for wealth. Finland already uses income-based traffic fines, assessing penalties known as “day fines” that are based on the offender’s daily disposable income.

“Personalized law … differentiates legal commands on the basis of relevant circumstances and thus avoids the inaccuracies and unfairness of crude uniform treatments,” the authors write. “This feature—higher precision—is the same reason that any personalized scheme is potentially superior to its one-size-fits-all alternatives. It is why custom-made shoes fit better than a single size, why personalized nutrition plans produce more effective diets than standard meal plans, and why medicine based on personalized diagnostics cures better than one-size-fits-all treatments. Uniformity, even if optimal on average, creates a poor fit for a lot of non-average people in a population with diverse preferences, characteristics, histories and means.”

Porat noted that some might wonder “how such a basic and simple idea has never been put to work," outside of a few notable exceptions that he and Ben-Shahar explore in the book.

"Maybe, some would think, it goes against the nature of the law, which must be uniform," Porat said. "Our thinking is that the personalization idea was not on the table until recently is because it was mostly not feasible: as in many other areas, law too had to wait for the data revolution to emerge.”

Avoiding the Pitfalls

Despite their optimism about the potential benefits of personalized law, the authors spend substantial time exploring the impediments: threats to equal protection if classifications like race, sex, religion, or gender identity allow for discriminatory impacts; the potential for data manipulation or other attempts to subvert the law’s intent; and the loss of social coordination. Uniform rules, after all, help smooth the flow of everything from traffic to trade to litigation.

Personalized law would also mean vast amounts of personal data in the hands of the government—a fact that might “raise a greater chill in our bones than so much data in the hands of Facebook and Google and Amazon and others,” Ben-Shahar said.

But, he added, we need to consider the tradeoffs and goals. Tailored speed limits, after all, would aim to do more than merely reward safe and experienced drivers—the object would be fewer accidents. Analysts would identify factors that make accidents statistically more likely—something insurance companies already do when setting automobile insurance premiums—and create algorithms designed around optimal safety. An inexperienced driver navigating a new city in the rain would be bound by a slower maximum speed than a safe and experienced driver navigating a familiar route on a sunny day—a calculation that would require government access to personal data but, ultimately, could save lives.  

“Yes, we value privacy, but at what cost?” Ben-Shahar said. “If we have 40,000 fatalities a year in traffic accidents, and data-driven regimes allowed us to reduce this by 30 percent, that's 12,000 lives.”

Algorithm design would be paramount, featuring much discussion about a personalized law’s specific goals and how to achieve those goals fairly. Existing data, of course, is flawed and often reflects deeply entrenched and longstanding bias: data reflecting discriminatory realities might, once used by algorithms, perpetuate these injustices, and potentially run afoul of equal protection laws.

Countering this would require further personalization, Ben-Shahar said, noting that individuals are more than their race, age, or gender; there are myriad factors that influence a person’s interaction with law and society. An algorithm that includes many points of data would minimize the effect of one factor and achieve greater precision. It would be more difficult to discriminate, even unintentionally, and it would be harder to game the system. And, most importantly for the law, people would no longer be classified according to salient but problematic suspect categories.

The conversation, Porat admits, has a “science fiction” feel to it, and cautions that “any implementation of personalized law has to be gradual, step by step, in areas where data about differences between people is particularly reliable and relevant.” He hopes that by putting this idea on the table,  the most essential questions about law and its goals can be revisited.

“We want readers to think differently about the question: ‘What is law?’” Ben-Shahar said. “Personalized law is a platform for people to rethink what criteria of equality really matter. [French writer] Anatole France famously chastised the sanctimony of equal laws, which forbid the rich and the poor to sleep under the bridges, to beg in the streets, and to steal bread. I hope that readers will see the benefits of personalized rules in specific contexts, especially important ones like criminal law, consumer protection, or road safety. Personalized rules could do a lot of good to weaker members of society for whom our so-called uniform laws haven’t been all that equal.”

He added that humans already intuitively understand that uniformity is “often unfair and counterproductive.”

“That’s why in many areas of our lives we treat people based on their personal characteristics,” he said. “Parents treat each child based on their needs and managers try to set personalized goals and incentives for their staff. Isn’t it time for Justitia, the Goddess of Justice, to remove her blindfold and treat people as individuals?”

Faculty books